Multi-hop Reasoning on Housing QA
82.67AccuracyTOTAL
Evaluation Results
| Method | Links | |
|---|---|---|
| TOTALModel=Claude, Prompting Method=TOTAL2025.10 | 82.67 | |
| CIC + COTModel=Claude, Prompting Method=CIC + COT2025.10 | 75 | |
| TOTALModel=Gemini, Prompting Method=TOTAL2025.10 | 74.33 | |
| CICModel=Claude, Prompting Method=CIC2025.10 | 71.67 | |
| CIC + COTModel=Gemini, Prompting Method=CIC + COT2025.10 | 70.33 | |
| TOTALModel=GPT, Prompting Method=TOTAL2025.10 | 70 | |
| CICModel=Gemini, Prompting Method=CIC2025.10 | 68.33 | |
| CIC + COTModel=GPT, Prompting Method=CIC + COT2025.10 | 66 | |
| CICModel=GPT, Prompting Method=CIC2025.10 | 64.33 | |
| COTModel=GPT, Prompting Method=COT2025.10 | 61.33 | |
| NAÏVEModel=Claude, Prompting Method=NAÏVE2025.10 | 60.33 | |
| NAÏVEModel=GPT, Prompting Method=NAÏVE2025.10 | 60.33 | |
| COTModel=Gemini, Prompting Method=COT2025.10 | 58.67 | |
| NAÏVEModel=Gemini, Prompting Method=NAÏVE2025.10 | 58 | |
| COTModel=Claude, Prompting Method=COT2025.10 | 57.67 |